|
|
Line Damages Simulation Generation into Synthetic Fingerprint
Jóriová, Vanessa ; Drahanský, Martin (oponent) ; Kanich, Ondřej (vedoucí práce)
Abstract The goal of this thesis is generation of line damage into synthetic fingerprint images. Creases, hair on sensor, and scars are specific damage chosen for generation. Attributes of all types of damage were analyzed in real fingerprint images and an algorithm for generation of this damage was proposed and implemented in Python. The quality of damaged fingerprints was evaluated by FiQiVi, the output of which is a percentage describing the quality of the fingerprint. VeriFinger was used to determine the effect on fingerprint matching, which is evaluated by a fingerprint matching score. Generation of damage into synthetic fingerprint images resulted in a decrease in both quality and fingerprint matching scores.
|
|
|
Analysis of Fingerprint Spoofs Created from Mold Done by Burning Technique
Sochová, Terézia ; Sakin, Martin (oponent) ; Kanich, Ondřej (vedoucí práce)
This work is devoted to the production and analysis of fingerprint spoofs created from mold using a burning technique. The aim of this work is to create fingerprints spoofs from different materials and then to design and to implement an algorithm that compares spoofs with model fingerprints. Spoofs are made using beech wood molds. In order to create the fingerprints, plasteline, silicone and liquid latex are used. The analysis of the differences between fingerprint spoofs and their real fingerprints is based on the minutiae extraction and on their subsequent processing. The results of analysis are represented by three monitored metrics: number and type of minutiae, number of paired minutiae and average distance difference between fingerprint spoof minutiae and real fingerprint minutiae. The created program is tested using a database of produced fingerprint spoofs. The results of the program are compared with the results of the VeriFinger application. Both programs agreed that spoofs made by silicone are the ones of the highest quality.
|
|
|
Generování onemocnění kůže do syntetických otisků prstů z Anguli
Míšová, Miroslava ; Kanich, Ondřej (oponent) ; Drahanský, Martin (vedoucí práce)
Bakalářská práce se zabývá onemocněním kůže postihující bříška prstů a dlaně. Pro tuto práci jsou vybrány nemoci dyshidrotický a hyperkeratotický ekzém. Otisky s těmito nemocemi z databáze výzkumné skupiny STRaDe jsou analyzovány. Je navržen postup, jak generovat tyto nemoci do otisků prstů vytvořených pomocí generátoru syntetických otisků prstů Anguli. Tento postup je následně implementován a vyhodnocen otestováním pomocí algoritmů na porovnání otisků prstů.
|
|
|
Advanced Generation of Damage Effects into Synthetic Fingerprints
Svoradová, Veronika ; Drahanský, Martin (oponent) ; Kanich, Ondřej (vedoucí práce)
The main goal of this thesis was to design and implement a application that would be able to generate fingerprint damage into a synthetic fingerprint. The application can create fingerprint images damaged by pressure, moisture and skin disease dyshidrosis with different intensity of damage. The application also allows annotation of the generated damage and its export. Selected damages were analyzed before the design was created. A database of fingerprints from five users was created to analyze the damage caused by pressure and moisture. The generated images and the achieved results are tested with VeriFinger and FiQiVi. For testing, 19 sets with fingerprints of different intensity and different type of damage were created. Experiments showed that the quality of the fingerprint decreased the most during the generation of moisture with the highest intensity of damage, where the quality decreased by 61.8 %. This thesis can be used for further research in the field of biometric fingerprint processing.
|
|
|
Simulation of Skin Diseases Effect Using GAN
Bak, Adam ; Goldmann, Tomáš (oponent) ; Kanich, Ondřej (vedoucí práce)
The aim of this master's thesis is to generate a dataset of synthetic fingerprint images that display symptoms of skin disease. The thesis deals with damage caused by skin disease in the fingerprint images and synthetic fingerprint generation. The diseased fingerprints are generated using a model based on Wasserstein GAN with gradient penalty. A unique diseased fingerprint database created at FIT BUT was used for training of the GAN model. The model was trained on three types of skin disease: atopic eczema, psoriasis vulgaris and dyshidrotic eczema. The generator network of the trained WGAN-GP model was used to generate datasets of synthetic fingerprint images. The synthetic images were compared with real fingerprint images using the NFIQ and FiQiVi quality assessment tools and by comparing minutiae location and minutiae orientation distributions in the fingerprint images.
|
|
|
Advanced Generation of Spoof Artefacts into Synthetic Fingerprints
Vrábľová, Žofia ; Drahanský, Martin (oponent) ; Kanich, Ondřej (vedoucí práce)
The goal of this thesis is to extend the application for spoof effects generation into synthetic fingerprints with the possibility of generation of two new spoof effects together with annotations of generated damages. Spoof effects chosen for this thesis are areas with lower clarity and defects in spoof material. Those effects were analyzed, methods to generate those effects were designed and then implemented. According to testing, generation of two new added spoof effects led to reduction in quality of fingerprint images, as well as the value of the similarity score determined during identification. In comparison with the original solution, the quality of the fingerprints decreased more in the extended solution, the similarity score in the generation of separate spoof effect decreased overall approximately equally.
|
|
|
Útok hrubou silou na přístupový systém s pomocí syntetických otisků prstů
Polehňa, Dominik ; Goldmann, Tomáš (oponent) ; Kanich, Ondřej (vedoucí práce)
Práce se zabývá odolností porovnávacího algoritmu VeriFinger vůči syntetickým otiskům. V práci jsou postupně rozebrány základy otisků prstů, porovnávacích algoritmů a generátorů syntetických otisků. Celkově byly navrženy tři experimenty, které měly pomocí syntetických otisků odolnost algoritmu vyzkoušet. Pro vyhodnocení jednotlivých experimentů byla implementována aplikace využívající algoritmus VeriFinger. Celkově bylo vygenerováno 2 800 000 syntetických otisků přes všechny experimenty. První experiment ukázal, že náhodné generování otisků není účinné v případě reálných otisků, avšak u syntetických otisků bylo do jisté míry dostačující. Druhý experiment dokázal, že lze pomoci porovnávání a skóre porovnání odhadnout třídu identifikovaného otisku a ve třetím experimentu se zúžením generování na jednu třídu zvedl celkový počet shodných otisků. Výsledky jednotlivých experimentů byly sepsány a navrženo jejich možné rozšíření.
|
|
|
Generování projevu falzifikátu do syntetických otisků prstů z generátoru SFinGe
Vrábľová, Žofia ; Drahanský, Martin (oponent) ; Kanich, Ondřej (vedoucí práce)
Cieľom práce je návrh a vytvorenie aplikácie pre generovanie prejavov falzifikátov do syntetických odtlačkov prstov z generátora SFinGe. Prejavy falzifikátov vybrané pre túto prácu sú prítomnosť vzduchových bublín, neprirodzený celkový tvar a náhle ukončenie papilárnych línií na okrajoch odtlačku. Vybrané prejavy boli analyzované, bol vytvorený návrh metód pre ich generovanie a metódy boli následne implementované. Generovaním implementovaných prejavov falzifikátov sa na základe testovania preukázateľne zhoršila kvalita syntetického odtlačku prsta. Pri vygenerovaní kombinácie všetkých implementovaných prejavov skóre v komerčnom nástroji kleslo priemerne o 49,37 % a kvalita odtlačku, určená metódou vytvorenou v rámci diplomovej práce pána Ing. Oravca, klesla priemerne o 6,18 %.
|
|
|
Synthetic Fingerprint Generation Using GAN
Dvořák, Jiří ; Drahanský, Martin (oponent) ; Kanich, Ondřej (vedoucí práce)
This thesis is focused on the generation of synthetic fingerprints using a model based on the principle of generative adversarial networks. The work summarizes the basic theoretical information about biometrics with emphasis on fingerprints. It also describes the principle of one of the popular synthetic fingerprint generators called SFinGe. The model based on a deep convolutional generative adversarial network is discussed together with several methods that improved its performance. The results were evaluated by computing the Fréchet Inception Distance between the generated and real fingerprints. The generated dataset of 100 samples was also evaluated by NFIQ 2.0 which proved that the proposed model is able to generate fingerprints with almost the same quality of the training samples.
|
|
|
Improvement of Methods for Detection and Classification of Damages in Fingerprint Images
Foltyn, Lukáš ; Heidari, Mona (oponent) ; Kanich, Ondřej (vedoucí práce)
This study aims to improve existing methods for detecting and classifying damage in fingerprint images by leveraging previous works conducted by students at Brno University of Technology. The work is built upon three applications: line damage (scars, hairs, creases) generator, moisture generator, and application containing multiple different models for fingerprint damage detection and classification. The three best-performing models - Faster-RCNN ResNet50, Faster-RCNN ResNet101, and CenterNet ResNet101 - were selected for further improvement. The work describes the creation of a dataset using undamaged synthetic fingerprint images, with the aforementioned damages introduced artificially. Efforts to improve the prediction accuracy of the models were based on more accurate annotation of bounding boxes and adjusting the hyperparameters. While the work yielded some improvements, the results are not consistently successful across all models and damage types.
|
host ::
RSS.